Homozygous MTAP deletion in primary human glioblastoma is not associated with elevation of methylthioadenosine

Abstract

Homozygous deletion of methylthioadenosine phosphorylase (MTAP) in cancers such as glioblastoma represents a potentially targetable vulnerability. Homozygous MTAP-deleted cell lines in culture show elevation of MTAP's substrate metabolite, methylthioadenosine (MTA). High levels of MTA inhibit protein arginine methyltransferase 5 (PRMT5), which sensitizes MTAP-deleted cells to PRMT5 and methionine adenosyltransferase 2A (MAT2A) inhibition. While this concept has been extensively corroborated in vitro, the clinical relevance relies on exhibiting significant MTA accumulation in human glioblastoma. In this work, using comprehensive metabolomic profiling, we show that MTA secreted by MTAP-deleted cells in vitro results in high levels of extracellular MTA. We further demonstrate that homozygous MTAP-deleted primary glioblastoma tumors do not significantly accumulate MTA in vivo due to metabolism of MTA by MTAP-expressing stroma. These findings highlight metabolic discrepancies between in vitro models and primary human tumors that must be considered when developing strategies for precision therapies targeting glioblastoma with homozygous MTAP deletion.

Fig. 1. Dramatic elevation of MTA in conditioned medium of MTAP-deleted cells versus WT coupled with mild intracellular elevation.

a The methionine salvage pathway (based on KEGG) with most genes are housekeeping expressed broadly across cell lines and tissues (DepMap, The Human Protein Atlas). SAM S-adenosylmethionine, dcSAM S-adenosylmethionineamine, MTA methylthioadenosine, SPD spermidine, PTN putrescine, MTOB 4-methylthio-2-oxobutanoate, DMTP 1,2-dihydroxy-5-(methylthio)pent-1-en-3-one, MTP 3-(methylthio)propanoate, MTDO-P 5-(methylthio)-2,3-dioxopentyl-phosphate, MTRi-P S-methyl-5-thio-d-ribulose-1-phosphate, 5-MTR-P methylthioribose-1-phosphate. b MTA levels were determined in conditioned media (extracellular) and washed cell pellets  (intracellular) of homozygous MTAP-deleted (MTAP−) and wild-type (MTAP+) cells in culture. Color coding reflects the same cell line for pellet/media. c MTAP status of cell lines confirmed by western blot repeated independently twice. d, e Levels of MTA (mean + SD, N = 3 biological replicates) in a panel of glioma cell; washed cell pellet—intracellular (d) and conditioned media—extracellular (e). Note >200-fold increase in MTA in conditioned media of MTAP-deleted versus intact cell lines (***p = 0.0005, unpaired two-tailed Student’s t test with unequal variance), while only marginal elevations are seen for this comparison in the cell pellet. f, g Time course of MTA levels in cell pellet and conditioned media (f) with SAM shown for comparison (g). x-axis: time of harvest after last media change; y-axis: ion counts for each metabolite. Each bar represents one biological replicate. There was a distinct, time-dependent increase in MTA in media of MTAP-deleted cells with a modest increase of MTA in the cell pellet. There was an imbalance of MTA in the pellet versus the media, contrasting what was observed with SAM, which is exclusively intracellular. h Absolute quantification of MTA in cell pellet and conditioned media (mean +/− SD, N = 3 biological replicates). The amount of MTA recovered from conditioned media of MTAP-deleted cells is 200-fold greater than recovered from the cell pellets. In contrast, in MTAP-WT cells, the amount of MTA in conditioned media is comparable to that in cell pellet. The modest intracellular accumulation of MTA (sixfold) in the MTAP-deleted cells compared to the WT cells is overshadowed by the increase of extracellular MTA. Significant values are indicated as ****p = 4 × 10⁻⁸, ***p = 0.00006, ^@@@@p = 10⁻¹², and ^%p = 0.02 using multiple t test with Bonferroni correction.

Fig. 2. No significant elevation of MTA and no specific inhibition of PRMT5 activity in homozygous MTAP-deleted primary resected GBM tumors.

a GBM tumors (heterogeneous mix of transformed glioma cells and non-malignant stroma) of defined MTAP-deletion status (genomic profiling from Kim et al.) were evaluated by mass spectrometry for MTA levels and western blot for SMDA. b Genomic copy-number data (dark blue: homozygous deletion, dark red: amplification) around the 9p21 locus. Each strip in the y-axis represents a single tumor at the specific chromosomal location (x-axis). c MTA levels in homozygous MTAP-deleted (MTAP−) versus MTAP-intact (MTAP+) GBM tumors, using the BIDMC mass spectrometric platform. Each bar represents MTA levels for each tumor (N = 1). There is a 1.4-fold increase in the median MTA levels (absolute ion counts) in MTAP-deleted tumors compared to intact tumors (p = 0.20, unpaired 2-tailed t test with unequal variance). d, e Same data but expressed normalized to total ion count for sample loading normalization (d) and as a ratio to SAM levels to account for methionine salvage pathway activity (e). Regardless of normalization, no significant elevation of MTA in MTAP-deleted tumors was evident. In contrast, f tumors with IDH1 mutation stand out by their dramatic elevation of 2-hydroxyglutarate (2-HG), providing a positive control for genomic/metabolic correlation. g To evaluate PRMT5 activity, GBM tumor lysates were immunoblotted for SDMA, repeated once. As a positive control, we assessed SDMA levels in MTAP-deleted U87 glioma cells in culture, alone or treated with exogenous MTA versus MTAP-reconstituted (U87-MTAP) cells. SDMA levels were lower in U87 than in U87-MTAP cells, indicating partial inhibition of PRMT5. SDMA levels further decreased following MTA treatment. Unlike in cultured cells, no specific decrease in SDMA levels was observed in MTAP-deleted versus intact GBM tumors. The presence of myeloid cells in a tumor is confirmed by the myeloid marker IBA1 in tumor lysates but not seen in cells in culture. h MTAP protein levels corrected for loading with GAPDH control (samples derive from the same experiment and gels were processed in parallel). Due to the presence of non-malignant MTAP-expressing cells in the lysate of whole tumors, MTAP protein levels are not zero for homozygous MTAP-deleted tumors; however, on average, they have lower MTAP protein levels compared to intact tumors.

Fig. 3. Primary GBM tumors are extensively infiltrated by MTAP-expressing non-malignant stromal cells.

a Immunohistochemistry (IHC) for MTAP and IBA1 on formalin-fixed paraffin-embedded (FFPE) sections of GBM tumors. b, c IHC was performed with monoclonal anti-MTAP. The slides were developed with EnzMet (black staining indicates MTAP expression) and counterstained with hematoxylin (blue, nuclei) and eosin (pink, cytosol, and extracellular space) on at least 50 cases stained with MTAP and IBA1 and representative of three independent cases are shown. Representative sections of MTAP-positive (b) and an MTAP-negative GBM tumor (c); x20 objective. Note the extensive MTAP-positive staining areas on the background of otherwise non-staining MTAP-negative glioma cells; histopathologic evaluation of morphology indicates that these MTAP-positive cells correspond to stromal components, including microglia, endothelial cells, and activated astrocytes. d Higher magnification view of c golden square. Any MTA secreted by MTAP-deleted glioma cells (blue dashed outline) stands to be phosphorylated by MTAP (producing 5-MTR-P) either released or taken up by MTAP-expressing stromal cells (white dashed outlines). e A representative section of an MTAP-negative GBM tumor stained against IBA1 antibody (myeloid marker), indicating the extensive presence of myeloid cells (microglia/macrophages) inside GBM tumors. f Higher-magnification view of e.

Fig. 4. Exogenous MTA is consumed by MTAP-intact cells through the methionine salvage pathway.

a MTAP-WT macrophages (RAW-264.7) were cultured with 100 µM methyl tri-deuterated-MTA (D3-MTA). Cells and conditioned media were extracted and prepared for LC-MS and NMR measurements, respectively. b Exogenous D3-MTA rapidly disappears from the media in cultures of macrophages, as the deuterium label (Mass + 3; M + 3) appears in intracellular methionine and SAM (c), consistent with MTAP-dependent metabolism by the methionine salvage pathway (d). Data in panel c are expressed as mean + SD, N = 3 biological replicates. e MTAP-deleted cancer cells accumulate MTA in conditioned media (detected by NMR) while secreted MTA is consumed by MTAP-intact macrophages, abrogating MTA accumulation in the extracellular environment. f Co-culture of MTAP-deleted cancer cell lines with MTAP-intact myeloid cells prevents loss of SDMA, indicating maintained PRMT5 activity, repeated once. MTAP-deleted and reconstituted glioma cells were co-cultured with myeloid leukemia cells (MV-4-11, a monomyelocytic suspension cell line). Before lysate preparation, suspended MV-4-11 myeloid cells were removed, and cancer cells were extensively washed. For each independent MTAP-deleted cell line, SDMA levels increased with myeloid co-culture, indicating restoration of PRMT5 activity.